Several years in and there is still quite a bit of confusion around the value of cloud computing. What is it? How can I use it? What value will it provide? There are several perspectives on how to approach cloud computing value. Interesting, that very question elicits several possible responses. This missive specifically targets how applications map against a cloud value matrix. From the application perspective, scale along with the historical component governs the direction of value.
As scale increases, so does the potential value from cloud computing. That is not to say that traditional methods are not valuable. It has more to do with the direction and velocity that the scale of an application is taking. Greenfield applications provide a different perspective from legacy applications. Rewriting legacy applications simply to use cloud brings questionable value. There may be extenuating circumstances to consider. However, those are not common.
Legacy vs. Greenfield (x-axis)
The x-axis represents the spectrum of applications from legacy to greenfield. Greenfield applications may include either brand new applications or rewritten legacy applications. Core, off the shelf applications may fall into either category. The current state of the cloud marketplace maturity suggests that any new or greenfield applications should consider cloud computing. That includes both PaaS and SaaS approaches.
The first step is to map the portfolio of applications against the grid. Each application type and scale is represented in relation to the others. This is a good exercise to 1) identify the complete portfolio of applications, 2) understand the current state and lifecycle and 3) develop a roadmap for application lifecycles. The roadmap can then become the playbook to support a cloud strategy.
The value cloud computing brings increases as application requirements move toward the upper-left quadrant. In most cases, applications will move horizontally to the right rather than vertically upward. The clear exception is web-scale applications. Most of those start in the lower-right quadrant and move vertically upward.
The matrix is intended to be a general guideline to characterize the majority, but not all applications and situations. In one example, legacy applications may be encapsulated to support cloud-based services as an alternative to rewriting.
Tim, I am looking forward to using this approach in my technology roadmap.